In talking about “methods” of training a certain behavior or chain of behaviors, we often get caught up in what will work for the dog we are training. Phrases like “I trained this dog with channel weaves; it worked better for him” or “this dog had to have more structure than a running dogwalk so I trained a stop” abound in agility training circles while in the behavior circles things can be a little more insidious; implying that some dogs require punishment tactics or harsh negative reinforcement procedures. In reality, while there are differences in personality, differences in motivation, and certainly differences in environment; all dogs do learn the same. The laws of learning are just that, laws, and we’d do well to recognize that if a solid training procedure fails for us it is unlikely that the dog is the reason why. So what are the reasons why? They are many, as you might have guessed, and I don’t claim to know them all. But the two major culprits I see are inadequate training skills for the task, and sloppy procedures.
When the Trainer is in Too Deep
So often it is our training skills that are failing us. We go from shaping scent discrimination to a tie down board, from 2x2s to channels, or from a functional differential reinforcement procedure to a bark collar; and we say it’s because the dog needed a different path. There is nothing wrong with throwing the towel in on a plan that isn’t working; in fact it’s a smart move. The fault lies in failing to recognize that it is us, the trainers, who are the shortcoming in the procedure. Sloppy markers, ill-chosen reinforcers, poorly arranged antecedents, and just plain lack of observation skills are good places to start if you’re not sure which of your skills could use work!
When the Roadmap is Faulty
Training plans are often at fault, of course. They need constant examining and actual planning to be effective. I find that faulty plans are everywhere in my sport of dog agility, and they continue to masquerade as trade secrets because they appear to be working so well for the people who market them. If you are pretty sure your training skills are not to blame, check these areas in your procedure for problems:
Rate of reinforcement–this is the biggie! Any solid training procedure will include a very high 1:1 (that’s one behavior for one reinforcer) reinforcement ratio. If the dog is going for long gaps of time without reinforcement during the initial acquisition of a behavior, something is awry. If you expect your dog to operate under a thin reinforcement ratio (because, perhaps, he has been sequencing a long time or is preparing for a show) and you are seeing problems, know that your rate is too low to sustain the behaviors you are expecting.
Type of reinforcement–we often jump to our favorite reinforcer without considering what that reinforcer does to our dogs in the learning environment. Toys and food, for instance, encourage very different emotional states. Just as we would not use boring cut-up carrots for a behavior that required speed and enthusiasm, we might not want to use the highly-exciting-life-or-death-stinky-udder-tug for behaviors that require thought and precision.
Antecedent arrangement–the environment has an effect on our work. We wish it didn’t but we do not train in a laboratory with perfectly controlled conditions. So, consider your antecedents. Is the barking dog outside messing with your dog’s ability to find the right scent article? Could the surface be contributing to your dog’s jumping capabilities? Are there competing motivators or information in your shaping environment that are detracting from the salience of your setup? When arranging the environment in which your dog will train, think of lighting the path to reinforcement; you don’t want your dog fumbling around in a dark closet searching for reinforcement!
So the next time you claim that your dog required a new plan, think about the other elements involved.